VFF - The signal in the noise
News

Bedrock AgentCore adds Chrome policies for controlled agent browsing

Read original
Share
Bedrock AgentCore adds Chrome policies for controlled agent browsing

Amazon Bedrock AgentCore now supports Chrome enterprise policies and custom root CA certificates, enabling organizations to enforce granular browser controls on AI agents. The feature allows configuration of over 450 browser settings including URL filtering, download restrictions, and password manager controls through familiar Chrome enterprise JSON configuration. Custom root CA support lets agents connect to internal services and work with corporate SSL-intercepting proxies. This addresses security risks from unrestricted web access and certificate validation failures on private infrastructure.

  • Bedrock AgentCore Browser now enforces Chrome enterprise policies with 450+ configurable settings
  • URL allowlists and denylists restrict agent navigation to approved domains independent of prompt engineering
  • Custom root CA certificates enable agents to connect to internal services and corporate proxies
  • Policies are managed at the browser level through the control plane API, separating security governance from agent development

As AI agents gain web access capabilities, unrestricted browsing creates security and compliance risks. This feature brings enterprise-grade browser controls to agent deployments, addressing a critical gap between agent flexibility and organizational security requirements. The separation of policy management from agent logic also enables security teams and development teams to operate independently.

Organizations deploying agents for sensitive tasks like invoice processing or data entry can now enforce browser restrictions without embedding policy logic into agent code. This reduces operational overhead, improves security posture, and enables faster iteration on agent capabilities while maintaining compliance boundaries.

  • Enterprise adoption of web-capable agents becomes more viable for regulated industries and organizations with strict security requirements
  • Security teams gain a familiar control surface through Chrome enterprise policies rather than requiring custom agent-level restrictions
  • Internal service integration becomes practical for agents, expanding use cases beyond public web access to proprietary systems and APIs

Monitor adoption patterns to see whether enterprises prioritize URL filtering, credential management controls, or certificate handling. Watch for emerging best practices around policy templates and whether AWS or third parties develop standardized policy libraries for common agent use cases. Also track whether other agent platforms adopt similar enterprise policy frameworks.

Share

Subscribe to the newsletter

The latest stories and analysis, delivered to your inbox.

Free. No spam. Unsubscribe any time.

Related stories

Warner Targets AI Agents in First Regulatory Framework
TrendingNews

Warner Targets AI Agents in First Regulatory Framework

Sen. Mark Warner plans to unveil a discussion draft bill focused on regulating AI agents, the autonomous systems driving much of the technology's current growth and spending. The bill aims to address emerging issues including user data confidentiality and whether large platforms like Google and Meta can restrict competing agents. This marks the first legislative framework attempt to create rules for AI agents, though passage faces headwinds from a crowded legislative calendar and midterm elections.

by Leo Schwartz· The Information
Cursor brings mobile app for remote coding agent control
TrendingNews

Cursor brings mobile app for remote coding agent control

Cursor has released a mobile app that enables remote management of coding agents. The app allows developers to oversee and guide AI-powered coding tools from mobile devices rather than being confined to desktop environments. This extends Cursor's existing coding agent capabilities to on-the-go workflows.

by Russell Brandom· TechCrunch AI
New agentic memory cuts token use 27x vs. competitors

New agentic memory cuts token use 27x vs. competitors

Researchers at the National University of Singapore developed MRAgent, a framework that dynamically reconstructs memory during reasoning rather than passively retrieving documents upfront. The approach significantly reduces token consumption and runtime costs compared to existing agentic memory systems, addressing a core limitation where context windows fill with irrelevant noise during long-horizon reasoning tasks.

by bendee983@gmail.com (Ben Dickson)· VentureBeat AI
Patronus AI raises $50M to stress-test AI agents

Patronus AI raises $50M to stress-test AI agents

Patronus AI, a startup founded by former Meta AI researchers, has raised $50 million to build digital worlds designed to stress-test AI agents. The funding round reflects strong investor confidence in the company's testing approach. According to its investors, the startup is experiencing nearly insatiable demand for its services.

by Marina Temkin· TechCrunch AI